Effect of exercise interventions in the early phase to improve physical function after hip fracture – A systematic review and meta-analysis

The efficacy of exercise interventions in the early recovery phase, i.e. started within the first three months after hip fracture, has been poorly studied compared to prolonged exercise interventions.To examine the effect of exercise interventions to improve physical function in the early phase after hip fracture.Seven databases including MEDLINE via Ovid, The Cochrane Library, Embase, Cinahl, Pedro, AMED and Web of Science were comprehensively searched till December 2019.Randomised controlled trials (RCTs) of exercise interventions initiated within the first three months after hip fracture to improve physical function, were eligible for inclusion. Primary outcome was physical function assessed using walking ability, walking speed, balance, muscle strength, mobility, and endurance.We conducted subgroup analyses specifically to investigate outcomes of these individual measurements. A meta-analysis was conducted to examine the overall effect of early exercise interventions. A meta-regression was conducted to examine the impact of study characteristic on exercise interventions. We used the PEDro score to determine quality of the included studies.Nine studies (669 patients) were included. Despite high statistical heterogeneity, there was high to moderate quality evidence that exercise provided benefit in improving physical function (standardised mean difference (SMD) 1.07; 95% CI: 0.44 to 1.70; p<0.001). There was no statistically significant difference in outcome, when measured by the individual physical function outcome (p>0.05). Meta-regression demonstrated no statistically significant association between study characteristics and exercise interventions (p>0.05).Exercise in the early phase of hip fracture rehabilitation can improve physical function. It remains unclear what type of exercise is superior in the early phase after hip fracture.This conclusion should be interpreted with caution given the high statistical heterogeneity reported and non-significant subgroup analyses of specific physical function measures, which were underpowered. Protocol Registration (PROSPERO): CRD42018091135.